Aleksey Voloschuk

[Aleksey Voloschuk @giwwi]

@danshipper @every This is not the end of automation. It may be the last wave of labor demand — the final expansion of human employment before agency, verification, and accountability themselves become partially automated.

[Aleksey Voloschuk @giwwi]

@iam_elias1 The problem isn’t whether a text was written by a human or AI. That’s the wrong focus.

What actually matters is verification — is the information accurate, and does it contain anything valuable? The method of creation is secondary.

AI detectors are basically a distraction.

[Aleksey Voloschuk @giwwi]

@jayvanbavel Very interesting. This gets at a deeper structural problem: when generating signals is cheap, small coordinated minorities can dominate attention and distort perceived reality. That is not just a media problem — it is a political economy of verification problem.

[Aleksey Voloschuk @giwwi]

@LuizaJarovsky This one feels personal.

I do not think the main issue is that AI creates fake competence out of nothing. The harder case is when the underlying idea is genuinely yours, but the writing, structuring, and refinement happen inside a human-model loop.

[Aleksey Voloschuk @giwwi]

@emollick Generation is already overwhelmingly in AI’s favor on cost grounds. Verification may follow.

If AI becomes far cheaper at both producing and evaluating knowledge, humans may end up less as authors or judges, and more as symbolic legitimizers of processes we no longer understand.

[Aleksey Voloschuk @giwwi]

My Substack currently resembles a graveyard of forgotten ideas.

The real question is: how much of what we bury is actually bad — and how much is merely unabsorbed?

New post: “The Graveyard of Unabsorbed Knowledge”

https://t.co/ilOPvvrjI6

#AI #Economics #PoliticalEconomy

[Aleksey Voloschuk @giwwi]

@tszzl knowledge diffusion ≠ benefit diffusion

[Aleksey Voloschuk @giwwi]

9/ The window in which a human can still verify text generated at machine speed is not permanent.
This book was written inside that window.
Full essay: https://t.co/GqFqILkCFX

[Aleksey Voloschuk @giwwi]

1/ My book took 15 years to think and 2.5 days to write down with AI.
That's not a flex. That's the theory in action. 🧵

[Aleksey Voloschuk @giwwi]

8/ I don't know if that's evidence for the theory or the most sophisticated self-deception available in 2026.
The full draft is open: 📄 https://t.co/ZFxZob2UEw 💻 https://t.co/YAEkb6WANm

[Aleksey Voloschuk @giwwi]

There are two futures for AI.

In one, it helps society distinguish signal from noise and turns information into knowledge faster.
In the other, it floods the world with cheap output while leaving verification as the central scarcity.

[Aleksey Voloschuk @giwwi]

6/
I wrote about this here:
https://t.co/rVW14SgMUe

[Aleksey Voloschuk @giwwi]

1/
One of the darkest things ever written about AI was written in 1947.

Not mass unemployment.
Not robot takeover.

A girl stops playing the violin.

[Aleksey Voloschuk @giwwi]

5/
A violin performance is not just correct notes.
Judgment is not just correct output.
Some human capacities only exist as biography.